<nettime> Design patterns, imposed developments and a fracture in Debian

[Jaromil <jaromil@dyne.org>]

For those interested in technopolitics: the concern around a necessary fork of
Debian is growing. This declaration generated some interesting threads on HN
http://debianfork.org . Pasting it here.

Needless to say I feel very much like the Veteran Unix Admins. Not debating if
this systemd is amazing good code or not - simply smells not to me, knowing
pulseaudio: really made by desktop minded people in comparison, for instance,
to jack

But srsly wearing the admin hat: I would never run anything so big as systemd
on my production servers before it has been at least 10 years around...


   ÂÂShall we fork Debianâ? :^|

   Â

    Who are you?!

   We are Veteran Unix Admins and we are concerned about what is happening to
   Debian GNU/Linux to the point of considering a fork of the project.

    And why would you do that?

   Some of us are upstream developers, some professional sysadmins: we are
   all concerned peers interacting with Debian and derivatives on a daily
   basis.

   We don't want to be forced to use systemd in substitution to the
   traditional UNIX sysvinit init, because systemd betrays the UNIX
   philosophy.

   We contemplate adopting more recent alternatives to sysvinit, but not
   those undermining the basic design principles of "do one thing and do it
   well" with a complex collection of dozens of tightly coupled binaries and
   opaque logs.

    Are there better solutions than forking?

   Yes: vote [1]Ian Jackson's proposal to preserve freedom of choice of init
   systems.

   Then make sure sysvinit stays the default for now, systemd can be
   optional.

   Debian leaders can go on evaluating more init systems, just not impose one
   that ignores the needs of most of its users.

    Why don't you do that yourselves?

   We are excluded from voting on the issue: only few of us have the time and
   patience to interact with Debian on a voluntary basis.

   Now we do what we can, hoping our concerns will be heard by those who can
   cast a vote about it.

 [edit/clarification]
 Since this seems to be one of the most prominent critiques, we'd like to clarify this point.
 With lack of time and patience we refer to our possibility to be involved in a complex
 bureaucratic system like the one governing Debian. While we respect this way of working,
 we think that our time is better invested in new directions, also according to our expertise.

    Is really all this fuss necessary?

   To quote Ian Jackson:

   "This resolution is not only important within Debian, and not only for
   jessie (its next release). It is also important feedback for upstreams,
   and our peer distros and downstreams".

    Why is this happening in your opinion?

   The current leadership of the project is heavily influenced by GNOME
   developers and too much inclined to consider desktop needs as crucial to
   the project, despite the fact that the majority of Debian users are
   tech-savvy system administrators.

    Can you articulate your critique to systemd?

   To paraphrase Eric S. Raymond on the issue, we see systemd being very
   prone to mission creep and bloat and likely to turn into a nasty hairball
   over the longer term.

   We like controlling the startup of the system with shell scripts that are
   readable, because readability grants a certain level of power and
   consciousness for those among us who are literate, and we believe that
   centralizing control services, sockets, devices, mounts, etc., all within
   one daemon is a slap in the face of the UNIX philosophy.

    How long are your beards?

   This is not a beard contest, rest assured the furry ones among us are not
   sheeps.

    To sum it up?

   If systemd will be substituting sysvinit in Debian, we will fork the
   project and create a new distro. We hope this won't be necessary, but we
   are well prepared for it.

    We need to talk.

   Sure, write an email to  VUA at debianfork dot org.

    Are you guys alone in this?

   Not at all, there are more protests against the imposition of systemd on
   users.

   This article is a good introduction to the issue at hand: [2]Systemd:
   Harbinger of the Linux apocalypse.

   There is the [3]boycott systemd website providing several references.

   Then there is the "systemd fork" called [4]uselessd with some good points
   and lots of lulz.

   The wikipedia page lists also some critiques in its [5]systemd reception
   section.

   With our protest we intend to represent the discontent of Debian users,
   because that's who we are. We intend to keep using Debian on our servers,
   or a fork if necessary. Others might have other goals, but we all share a
   common problem: systemd being imposed on us.

    Thanks for doing this. How can I help?

   Cheers.

   You can help by talking to fellow Debian developers and convince them of
   how wrong is to betray a very big and relevant userbase by listening to
   desktop needs.

   Also it can be helpful to monitor and update the [6]Wikipedia page about
   systemd.

    How are people reacting to your protest?

   Here below some of the messages we are receiving. If you write us please
   specify if you allow us to quote your nick, else we may quote you
   anonymously, or not quote you at all if you specify so.

    I'm with you guys in the way that systemd is not the way forward. But I do
    not think that sysvinit is the way forward either. Maybe another init
    system? Like GNU DMD? I know its still pretty immature, but I think it's a
    good init system. I don't think sysvinit is ever going to make a comeback.

    And forking Debian? That's a very hard feat. Debian is the biggest Linux
    distro to date with hundreds of developers and at least 10 times as much
    users. And you'll have to change a lot of things from upstream. Even
    Ubuntu decided to use systemd instead of having to change the base init to
    upstart. Perhaps going to Slackware instead? That's what I use. Its very
    UNIX like. If that's what you strive for.

 Thank you so much for this. I've been using Debian since Hamm and this
 systemd nonsense has me ready to jump ship.

    I don't know who is behind that email, but sincerely, thank you for
    doing this.  I'm an UNIX/Linux sysadmin for nearly 20 years, I am
    nowadays dealing with a 5k servers which consists of nearny 90%
    debian systems. I've been a long time opponent to systemd, first
    because I read the code (thing that too few of that crap's zealots
    do), and ultimately because I tried it. That thing is a desktop
    toy, and even then, it has failed me on 50% of the cases. Its very
    nature is an abomination to UNIX principles and me and my team,
    colleagues, friends on the sysadmin field are *VERY* worried (to
    say the least) on what's coming.  Please keep that movement going,
    make it strong, and if you want a hand, count me in.

   Again, thank you.

 I've been using runit as my init system on debian wheezy/jessie for a
 while now and it works pretty well. glad to see the effort though on
 debianfork.org . cool.

    Do you realize that there's already more than 150 Debian
    derivative?  Instead of doing one more, you're better off helping
    one of the already existing derivatives. Writing "fork" is the joke
    part, since that would only be a derivative, unless every Debian
    Developer follows you, which will *not* happen.

    Also, instead of just writing words on the internet, wouldn't you
    think that helping some of the systemd alternative be a lot more
    productive?

    FYI, I don't like systemd either. And I've been maintaining OpenRC
    in Debian, and trying to push for it to work on as many arch as
    possible.  Though I currently don't have the time for it (for
    professional and personal reasons), and some others are a bit
    taking over the work, but it's not going as fast as it should. Some
    help would be awesome, and would help a way better than writing
    funny text on the net.

 I will support you guys with code if the fork goes ahead. But kindly
 think of another name.

    Why don't you guys help out with Debian LTS? No systemd there, and
    they need help at the moment.

 I'm in, at least in general.  "because readability grants a certain
 level of power and consciousness for those among us who are literate"

 I'd edit that to add that systemd is for software devs, but sysvinit
 is for sysadmins. Sysadmins can debug shell scripts, but not
 necessarily debug and recompile systemd, which will of course be
 necessary in the course of admin duties. Excessive product tying and
 complication for the sake of complication takes away the ability of
 admins to administer, which is highly counter productive. The
 response from the systemd astroturfers is invariably along the lines
 of software, especially our software, will never have bugs (LOL)

 "Pure Debian by Veteran Unix Admins." any other name has to be better
 than that. At least try not to violate trademarks by putting in (tm).

 You missed an excellent lack of freedom argument. There will be only
 one way to do it and it will be mandatory and all else will be
 forbidden and only one guy gets to decide. Intentional
 incompatibility is a pretty screwed up way to go thru life. It does
 not help that most of the new features and abilities of systemd appear
 totally useless. If one mp3 player was useless and awful, that would
 be OK, unless the inquisition extinguished all other mp3
 players... that would suck. Although this is whats being done with
 systemd.

 And the embrace extend extinguish issue. If submarine software
 patents sink apache (unlikely, but possible) then I don't really care
 although the emergency conversion to nginx might be a PITA for a day.Â
 When the one and only init system to bind them all is subverted either
 by submarine patent or security holes then the temporary work around
 will be to install the xyz package to switch to ... well I guess it'll
 be impossible and I'd have to switch to freebsd.

 There is a cost - benefit ratio issue. Most of what systemd is
 capable of doing is unfortunately completely useless and irrelevant in
 comparison to what is being lost WRT to ease of debugging and
 reliability and security and freedom. If the costs were low/zero, the
 useless features it provides would likely still not be worth it, but
 at least the ratio would make it less awful.  To some extent the whole
 situation is a farce. You've got software devs deciding what desktop
 users that don't exist want, while ignoring actual current desktop
 users, and they're re-applying that "successful" (LOL) development
 model to screw over sysadmins, and we're not subject to being told
 what we want by non-admins. Its a very arrogant business model.

    If/when you start publishing work for your fork

    Can you see if it can be documented in a way those of us in the redhat
    world could take advantage of it? We have the same problem and concerns,
    well at least the could *nix admin in our shop. Myself included.

    We are also considering looking at alternatives and we all have used
    debian in the past and most use ubuntu for desktop (there is always that
    one mandrake guy) so I will be keeping an eye on this project as well.

 Whatever the outcome you have my support !

 Using various Unices since System V on an ICL Clan 4, and as long-time
 BSD, Slackware and Debian user, I just donât want systemd imposed on
 me, whatever the purpose. Whether itâs total crap, a good idea or a
 mix of both, this has to be *optional* and surely not a default. I
 want to keep my servers and embedded systems bloat-free with only
 their intended software sets installed.  Why donât *they* fork Debian
 and do whatever they want ? I like Debian the way it is (was) like a
 good grilled steak with nothing superfluous added, no fancy
 sauce. Thank you for your efforts !

    I've been on Linux for over 15 years, I'm also a contributor in
    Fedora and author of NetworkManager-ssh (SSH plugin for Network
    manager). I'm a sort of a cross between a sysadmin and developer,
    currently doing both.

    I'm predominantly a Fedora user (and contributor) and am thinking
    recently about forsaking Fedora in favour of something without
    systemd (was thinking PCBSD). I've been around Linux for a while
    and understand the core philosophy behind it and how systemd just
    betrays all of that. I actually had an argument with Mr. Pottering
    on the Fedora-dev list about having /var/log/messages having a
    binary format. When log files in linux will not be plain text -
    it'll be the end in my opinion. needless to say Pottering didn't
    agree with me and tried to shut me up.

    Pottering somehow managed to push systemd quite far in Fedora and
    at the moment I'm in absorption mode trying to digest WTF happened
    to my Fedora with systemd. Horrow show.

    If this fork goes forward, I promise to try and do my best to
    contriube as much as possible. I'm good with packaging and shell, I
    believe I will be able to help. I think you guys hit the nail on
    the head with your web page and you have my full support.

    Lets not let our voice be silenced

 As a veteran (20+ years; Solaris, HP-UX, RHEL, Debian) sysadmin who
 prefers Debian over everything for use on about 40 GUIless servers,
 it's either this (a fork away from systemd) or I move away from Debian
 altogether.

    Many thanks for the effort here, while it's not a big thing to
    replace systemd on jessie/sid installation with sysvinit, it's not
    easy to create an installer image which installs the OS without
    systemd. I made an image which you can find on
    http://without-systemd.org/debian-jessie/ The big problem is
    tasksel, tasksell will install all Packages with the priorities
    required, important and standard. Unfortunately systemd has the
    priority importand and will be installed again, if you select
    "standard system utilities" in tasksel, to fix that, i would have
    to create my own Debian repository :-(

    I hope for a solution without a fork. But if you need help with
    forking, please contact me.

 I support your idea of forking Debian to exclude systemd and its
 multitude of tentacles. Please follow through. Also, please make
 available means by which us mere users (though I've got kernel-hacking
 and systems-architecture ambitions myself) may participate in an
 effective way.

 As I see it, the simplest way to fork Debian as of october 2014 would
 entail taking a full-system snapshot (per package versions), available
 from snapshot.d.o, from a time before systemd dependencies were
 introduced, and then adding on top any updates that didn't add
 such. This would re-use nearly all mainline Debian effort and require
 only porting of e.g. eudev from other systemd-less distributions. Of
 course GNOME would have to go; fortunately there is MATE.

    Thanks for doing this. I run four Debian servers in production,
    three of which are connected to an IRC network. Gnome remains the
    default DE on Debian for accessibility reasons, so it's obvious
    they have the monopoly of votes in favor of systemd. Large
    companies like Red Hat are also backing destroying the
    UNIX-philosophies.

    Once I heard Debian would be making the switch to systemd, I've
    been slowly migrating all Debian servers over to OpenBSD over the
    impending death of Debian. There's still much that leaves me
    missing Debian, most importantly the large amount of different
    packages that are not available in OpenBSD and would take lots of
    effort to port over.

 I have been worried about systemd colonizing all the linux distros for
 a long time and it's really happening. I want users to retain freedom
 to use the init system they choose rather than being locked in.

 The debian mailing list vote about whether to switch to systemd was
 atrocious, in no way did it come to a consensus but they still forced
 this upon us? This was the reason I quit using debian. If you look
 around there are very few remaining distros that aren't getting taken
 over by systemd!

 Your project is a good idea and I wish you the best.





     *ÂAll information on this page is free to copy. This webpage is an
       independent communication promoted and managed by a group of Debian
       users, developers and admins and is not affiliated with the Debian
       project. Debian is a registered trademark of Software in the Public
       Interest, Inc.

References

   Visible links
   1. https://lists.debian.org/debian-vote/2014/10/msg00001.html
   2. http://www.infoworld.com/article/2608798/data-center/systemd--harbinger-of-the-linux-apocalypse.html
   3. http://boycottsystemd.org/
   4. http://uselessd.darknedgy.net/
   5. http://en.m.wikipedia.org/wiki/Systemd#Reception
   6. http://en.m.wikipedia.org/wiki/Systemd



-- 
Jaromil, Dyne.org Free Software Foundry (est. 2000)
We are free to share code and we code to share freedom
Web: https://j.dyne.org Contact: https://j.dyne.org/c.vcf
GPG: 6113 D89C A825 C5CE DD02  C872 73B3 5DA5 4ACB 7D10


#  distributed via <nettime>: no commercial use without permission
#  <nettime>  is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: http://mx.kein.org/mailman/listinfo/nettime-l
#  archive: http://www.nettime.org contact: nettime@kein.org